PM-201a · Module 1

Persona Engineering

3 min read

Persona engineering is the practice of shaping the model's operating assumptions about itself through the role definition. An expert persona activates different knowledge and communication patterns than a peer persona, which activates different patterns than a service persona. These distinctions are not cosmetic. They change what vocabulary is used, what level of detail is assumed in the audience, what tone is adopted, and how assertively the model makes recommendations.

  1. Expert persona "You are a senior cybersecurity architect with 20 years of experience." The model produces authoritative, technically precise output. It makes recommendations. It identifies risks. It assumes the audience is less expert than the persona. Use this when you need domain authority and directness in the output.
  2. Peer persona "You are a software engineer on our team working through this problem collaboratively." The model produces collaborative, exploratory output. It suggests and discusses rather than directs. It assumes equivalent expertise in the audience. Use this when you need a thinking-partner interaction rather than an authoritative deliverable.
  3. Service persona "You are a customer success specialist helping a first-time user." The model produces patient, explanatory, accessible output. It does not assume prior knowledge. It anticipates confusion. Use this when the audience is less expert and the goal is clarity over authority.

Persona engineering extends beyond expertise level. You can shape communication style within a persona: formal versus conversational, verbose versus terse, recommendation-making versus options-presenting. A senior attorney persona that also communicates in plain English for a non-legal audience is a compound specification — and it is a legitimate one. Combine persona dimensions deliberately. Don't leave any dimension to default.